Innovation Ability in Palaeognath Birds

Innovation reflects flexible cognition and is widely used as a general indicator of intelligence. In birds, innovative behaviour has been studied extensively within the more diverse clade Neognathae, particularly among parrots and corvids, and this research has produced a large body of literature. In contrast, the cognitive abilities of palaeognath birds have received far less attention. A recent study addressed this gap by designing an experiment to test innovation in nine individuals representing three palaeognath species.

Modern birds are divided into two major lineages: Neognathae and Paleognathae. Most living bird species belong to Neognathae, including familiar birds such as sparrows, crows, chickens, parrots, and pigeons. Paleognaths, by comparison, consist largely of large flightless birds such as ostriches, cassowaries, and emus. The study tested four common ostriches (Struthio camelus), two greater rheas (Rhea americana), and three emus (Dromaius novaehollandiae).

Common ostrich（圖片來源：Yathin S Krishnappa，採用 CC BY-SA 4.0 授權。）

The experiment used a specially designed device consisting of a plastic rotating disk fixed with a bolt. Behind the disk were several small chambers, some containing food. The disk had a single hole; when the hole was rotated into alignment with a chamber, the bird could access the food inside. Before testing began, the birds were given a 30-minute familiarization period during which they encountered the apparatus without the rotating disk attached. During this phase all emus approached and interacted with the device, whereas the rheas and ostriches did not approach it.

Left: the experimental apparatus; right: greater rhea R1 manipulating the device（圖片來源：Clark, F. E et al. (2025). ，採用 CC BY 4.0 授權。）

During the formal experiment, all three emus and one rhea (individual R1) discovered a solution during the first problem-solving session. They rotated the disk by pecking or biting near the hole, aligning it with a food chamber and obtaining the reward. Across the trials this behaviour occurred 52 times: 42 by emus and 10 by the rhea.

The rhea R1 also demonstrated a second method of obtaining food. It bit the head of the central bolt and repeatedly twisted it from side to side, eventually loosening and removing the bolt. Once the bolt came free, the disk detached from the base and all food chambers were exposed simultaneously. This occurred twice during the experiment. However, after R1 discovered the simpler method of rotating the disk to obtain food, it stopped removing the bolt altogether.

Traditionally, birds with relatively large brains—such as crows and parrots—have dominated research on avian cognition. More recently, attention has begun shifting toward species with smaller brains and rarer taxonomic representation. Palaeognaths were often assumed to have extremely small brains and therefore were considered poor candidates for cognitive research. Consequently, relatively few studies had examined their cognitive abilities. The present study demonstrates that at least some palaeognath birds can employ physical manipulation strategies to obtain food, contributing important new evidence regarding their cognitive capacities.

The experimental apparatus differed from traditional puzzle-box tasks commonly used in animal cognition research. In conventional designs, animals must remove an obstacle to access food, representing a subtractive action. In this study, the task required aligning a hole with a food chamber by rotating the disk, representing an additive action. These two types of problem-solving may involve different cognitive processes, and future experiments comparing additive and subtractive tasks across bird species may provide deeper insights into animal cognition.

Wild emu photographed in the Tidbinbilla Nature Reserve, Australia（圖片來源：JJ Harrison，採用 CC BY-SA 4.0 授權。）

The researchers categorized innovation into two general forms. One involves using existing behaviours in new contexts, while the other involves creating entirely new behaviours. The wheel-turning solution used by emus and the rhea falls into the first category because pecking and biting are already common behaviours used for feeding, exploration, and feather maintenance.

The bolt-removal behaviour displayed by R1 represents the second category. Initially, the behaviour may not have been directed toward obtaining food; the shiny bolt might simply have attracted the bird’s attention. Once the action accidentally produced access to food, the behaviour was repeated. Nevertheless, once the bird discovered the easier solution of rotating the disk, the bolt-removal behaviour disappeared from its repertoire, suggesting that animals may replace complex solutions with simpler ones when possible.

The ostriches did not successfully solve the task. Researchers proposed several possible explanations, including lower motivation, fear of novel objects, difficulties manipulating the apparatus due to its height relative to their body size, or potentially lower cognitive ability associated with their smaller relative brain size.

Although the study revealed substantial behavioural variation among individuals and species, the sample size was small, meaning the conclusions remain preliminary. More extensive experiments will be required to confirm these findings. Nonetheless, because palaeognath birds occupy a relatively basal position in avian evolution, studying their cognition may also provide clues about the cognitive capacities of non-avian dinosaurs.

Author: Bai Leng

Reference:

Clark, F. E., Burdass, J., Kavanagh, A., King, A. (2025). Palaeognath birds innovate to solve a novel foraging problem. Scientific Reports.